The topics described in this chapter span an enormous variety and are too numerous to discuss in any detail here. Instead, we briefly describe a few sample applications and their corresponding data requirements. In many cases, the changes of interest are quite subtle or take place over very long times, and SAR interferometry may not be sensitive to them. On time scales of years to decades, meter-level or greater changes in topography may occur, but these are not the primary "observable" for many of the studies described here. Of far greater importance is an accurate, high-resolution description of the existing surface character and topography. Thus, SAR has an important role to play by generating high-resolution all-weather images, while the critical role of SAR interferometry is generating high-resolution topographic data.
Together, elevation, slope, and aspect (north vs. south facing) exert enormous control on surface hydrology and ecosystems. Especially in high-relief terrain, topography influences intercepted solar radiation, precipitation and runoff, evaporation, snow ablation, soil moisture, and vegetation type and health. Even in low-relief terrain, topography has signficant influence on soil moisture. Energy and mass-flux conditions determine the dominant types of vegetation and their succession, and these fluxes are affected by topography through combined influences of elevation, slope, aspect, and roughness. Calculation of accurate short- and long-wavelength exchanges in mountainous areas requires accurate information on neighborhood topography, including distribution of local horizons for shading, scattering, and reradiation estimates. Topographic parameters determine the exposure of a landscape to weather and sunlight at a given latitude and, thus, determine its microclimate. By feedback mechanisms, vegetation itself influences energy and mass fluxes, affecting not only local environment but also regional and global climate. Thus, topography is a key element in the study of complex ecosystems.
Existing topographic data are deficient in one or more respects for many of the applications described in this chapter. This is especially true outside of North America, western Europe, and Australia, but even high-resolution digital topographic data within North America may have tens of meters of error (Pike and Thelin, 1989; Fig. 2.2), causing problems with the differencing calculations necessary to derive slope and aspect. For large regional and global studies, the problem is compounded by the different vertical--and to a lesser extent horizontal--data used, as well as differences in the original map data from which most digital data are generated (e.g., different map projections, styles, and accuracies). This issue has been covered by several recent reports (e.g., Burke and Dixon, 1988; Mueller and Zerbini, 1989; Rundle, 1990) and we will not reiterate it here, save to update or refine some of the requirements for topographic data for the specific studies discussed below.
We also note that the Global Topography Mission, currently under study by NASA, would alleviate the major deficiencies in the global topographic data base, and we endorse that mission. In the meantime, high-resolution topographic data could be generated in some areas using data from ERS-1 and ERS-2, as well as NASA's TOPSAR aircraft, all of which should be pursued. Release of topographic data currently embargoed by the U.S. Department of Defense (DOD) should also be vigorously pursued. Although the accuracy of such data is often exaggerated, it is all that is available in many regions. Since the current plan for the Global Topography Mission includes the use of a laser altimeter, we have called out applications that would benefit from those data as well where appropriate.
Basin and Landform Evolution
Landforms and stream-channel network geometry vary over a broad range of spatial scales. Space-based topographic data will enable us to acquire a basic description of features with broad spatial scales, such as entire drainage networks, that have been poorly served by the available data. The ability to accurately detect changes in topography will be especially useful for studying geomorphic processes, such as progressive sedimentation or trenching of alluvial fans, broad floodplain sedimentation patterns and delta exten-sion, formation and movement of large dune fields, and large-scale development of gullies--e.g., the type rapidly forming in steep, intensively cultivated terrains in Asia, Africa, and South America.
The combination of landforms and stream networks in any region partially controls hydrologic mass fluxes, such as water and sediment discharge in stream and rivers. What has not yet been achieved, but is possible in principle with accurate space-based data, is a complete mass balance of fluxes of both water and sediment through a drainage basin. High-accuracy topographic and topographic change data are the keys to assessing the sediment component. The extent of vertical lowering (related to the volume of material removed) can be estimated with accurate topographic change data monitored over years to decades. Flow paths through the watershed can be accurately mapped from high-resolution digital topographic data, while ancillary measurements of stream sediment transported into the ocean complete the mass balance. In this way, subcontinental-scale erosion and depositional cascade can be assessed.
The physical variables that determine runoff and sediment transport include climate, vegetation, soil, and slope. Landforms and channel network geometry, readily calculated from DEM data (Marks et al., 1984) contain fundamental signatures of these parameters. Paleohydrologic analyses of river channels can provide indicators of past climates and, therefore, of climate change (Bull, 1991). Most of the empirical research on this topic has been limited to small basins because of the extensive field work required. As the use of high-resolution satel-lite imagery becomes widespread, theoretical advances are possible, including spatial analyses of channels and slopes . The scaling invariance observed in net-work geometries appears to reflect peak flows in the channels and sediment flows that connect channel slopes and peak flows. Careful tests of these theories will require a large data set from different climates and spatial scales. The combined topography and surface change aspects of SAR interferometry, especially a space-based system, offer the spatial coverage needed as well as the opportunity to get direct measurements of a system in which a peak flow has recently occurred. These data can test our ability to see the long-term changes in network geometry that may result from climate change. They can also provide scientists with before-and-after images of peak discharge events in stream geometry to help define those changes and the stability of those changes over periods of several years after peak discharge.
Requirements. Horizontal: 10 to 100 m; Vertical: 5 cm to 5 m;
Sampling: 5 days to annually; Rate of change: days to years.
Stream Channel Evolution, Basin Flooding
Stream channels respond on short time scales to regional events, such as changes in discharge during flooding, and to local changes, including increased runoff from urbanization. Thus stream-channel geometry is a sensitive indicator of change within a watershed. As areas become increasingly developed, the rate and volume of runoff increases since infiltration is reduced and fewer plants are available to store and use the surface water. This increase in runoff results in faster occurring, more frequent, and more extensive flooding. In turn, this change in stream dynamics produces an enlargement of the stream channel because of increased erosion of the channel during times of high discharge and lesser amounts of in-channel deposition between flood events.
Predictions for the Midwestern floods during the summer of 1993 were based on a detailed knowledge of that area's hydrological conditions and near-stream topography. For most of the Earth, such predictions are not possible because key hydrologic parameters are not known and cannot be determined because of logis-tical or financial limitations. A combination of SAR images and laser altimeter data should allow measurement of flood areas and river heights, critical parameters for short-term flood assessment as well as longer term understanding. During the flight of any single spacecraft, there will probably be extreme, peak discharge events on at least a small percentage of the world's major rivers, providing a short-term assessment of changes caused by increased or decreased flow, as well as intermediate-term recovery of the system towards an equilibrium state. Regardless of local weather, a series of SAR images may itself be a great benefit, allowing observation of stream channel changes on a frequent basis during flooding events.
Requirements. Horizontal: 10 to 100 m; Vertical: 5 cm to 1 m;
Sampling: daily to annually; Rate of change: days to years.
Soil Moisture
Soil moisture controls plant growth, the hydrological behavior of the soil, and the ability of the soil to resist erosion. Lower amounts of humus and plant root structure in areas of sparse vegetation make these soils especially vulnerable to wind erosion during periods of lower soil moisture. The Sahel region in Africa illustrates the sensitivity to climatic variations, ground water withdrawal, and overgrazing. Seasonal changes in the Sahel reflect the changing availability of water and show the effects of both short- and long-term changes in the near-surface water budget. Signifi-cant areas of the western U. S. are characterized by semiarid conditions and show a similar vulnerability to reduced soil moisture and increased erosion.
Soil moisture is strongly influenced by local topography, and the predilection for saturation following rainfall events can be easily calculated from high-resolution digital topographic data (O'Loughlin, 1986). The amount of moisture stored in the upper soil layer changes the dielectric constant of the material and thus may affect the SAR return. In fact, there is some evidence that SAR can be used to monitor changes in soil moisture, at least in the uppermost horizon. The spatial and temporal coverage permitted by satellites makes this a promising approach for measuring short-term, seasonal, and long-term variations in soil moisture and for determining how these variations affect primary productivity and erosion rates.
Requirements. Horizontal: 10 to 30 m; Vertical: 2 to 10 cm;
Sampling: 5 days to annually; Rate of change: days to months.
Canopy Structure, Seasonal Changes, and Logging Impacts
Forests are crucial sinks and stores of carbon dioxide (CO2) and provide a habitat for a wide variety of species. Forests worldwide are being destroyed or reduced by clear-cutting or selective cutting. As a response to resulting changes in sunlight and available nutrients, forested areas and plant species within them experience changes in growth patterns and species distribution. Secondary growth is vitally important to the long-term viability of these forests because of its role in controlling soil erosion and related loss of nutrients and its water absorption capability.
A laser altimeter is a possible complement to any space-based SAR system. Together they can acquire data that could prove very useful for vegetation monitoring. The structure of the return laser altimeter pulse is partly a result of the reflectivity of plants near the canopy top (Fig. 5.1a, Fig 5.1b, and Fig 5.1c). Combined laser altimetry and SAR imagery and interferometry can show areas of clear-cutting as well as estimates of the rates of regrowth where extensive logging has occurred. As an example, the flanks of Mt. St. Helens have been logged repeatedly, and areas of newest cutting and regrowth have been detected with this approach. A satellite-based system incorporating this technology could provide global information on the gross style of vegetation cover and, perhaps more importantly, changes in that cover.
Requirements. Horizontal: 10 to 100 m; Vertical: 10 cm to 5 m;
Sampling: weekly to annually; Rate of change: weeks to years.
Crustal Motion from Solid Earth Tides, Atmospheric Fluctuations,
and Tidal Loading
Both the solid Earth and the oceans respond to gravity perturbations caused by the tidal potential of the Sun and the Moon. For the solid Earth, there is a well understood elastic component (the solid Earth tide), and a less well understood component due to ocean-tidal loading of near-shore areas. Ocean tides are reasonably well modeled by the laws of hydrodynamics (Parker, 1991). These directly load the crust and induce a second-order gravitational potential perturbation at the same tidal frequency. The total effect of ocean-tide loading produces distortions in the crust of up to 15 cm, especially over continental areas near seas or oceans with strong ocean tides, such as the European continental shelf (Fig. 5.2) and Patagonia. The details of ocean-tide loading, especially in areas of complex tidal regimes, are not well known at present. Typical wavelengths associated with tidal loading range from 50 to 2000 km. Shorter wavelength response can also be expected near resonant seas or bays (e.g., the Bay of Fundy, Gulf of Tonkin). The signal magnitude, combined with moderate wave-lengths, suggests that this phenomenon should be detectable with SAR interferometry.
Fig. 5.1b. A time series of laser altimeter data for a partially vegetated slope.
Fluctuations in the atmosphere also load the solid Earth. Though the corresponding spectrum is continuous (from a few hours to several months), the response mechanism of the solid Earth is the same as for tidal loading. The vertical ranges are a few centimeters, with typical scales corresponding to the structures of the atmospheric pressure high and lows (100 to 1000 km) (see references in Kakkuri, 1991, and VanDam et al., 1994).
These various crustal deformations are very difficult to separate from each other using conventional observing systems. The tremendous spatial sam-pling offered by differential SAR interferometry would permit true improve-ment, rather than mere confirmation, of existing models.
Requirements. Horizontal: 10 to 100 m; Vertical: 10 to 30 cm;
Sampling: weekly to a few times annually; Rate of change:
periodic on many time scales.
Shoreline and Estuarine Processes
Coastal areas are a small percentage of total continental area, but are very important economically. We focus on two processes where high-resolution topographic data and SAR imagery have an important role to play, namely coastal erosion and the biological and geological processes taking place in tidal and estuarine areas.
Near-shore waves and currents redistribute sand and other materials along coastlines. These processes change the shore on a variety of time scales. Storm surges are an important influence on the near-shore sand budget and can cause large and rapid changes in shoreline. Every year, storm surges remove thousands of tons of sand over one to several days from beaches along the east coast of the U. S. More gradual changes over years to decades are also effective at modifying the shoreline. Seasonal changes in beach geometry are well docu-mented, while longer term erosional and depositional cycles are less well understood. All these processes remove material from beaches and deposit it both offshore and in navigable inlets. Short- and long-term monitoring of shorelines would allow a better understanding of the complex processes shaping this economically important interface between land and ocean. Even repeated SAR images by themselves could be useful for studying these processes.
Near-shore topography is important in studies of shoreline biological and geological processes, especially in estuaries and wetlands. The water or soil moisture available to support biological activity is affected not only by the tidal changes in water level, but also by the microtopography, which can change on time scales of days to months. High spatial resolution is required to map the subtle variations that influence the biologic and geologic processes of interest.
Requirements. Horizontal: 10 to 30 m; Vertical: 5 to 30 cm;
Sampling: 5 days to annually; Rate of change: days to months.
Subsidence
Many major cities are located in areas undergoing subsidence as a result of withdrawal of ground water, oil or gas, and other minerals. Several meters of subsidence over several decades are not uncommon. Examples of cities with significant problems include Houston, Mexico City, Maracaibo, and Katmandu. High rates of subsidence can have a major impact on flood control, utility distribution, and water supply. Subsidence is also a result of natural processes such as limestone or marble dissolution that forms karst topography. In western Pennsylvania, an underground coal fire has burned for many years causing localized subsidence and threatening a much wider region with similar risks. Successive SAR images in urban areas over periods of several months may be able to detect subsidence directly. The surface structure of many parts of urban areas remains unchanged over several years, suggesting that interferometry over several years may be possible. Subsidence rates of several centimeters per year or more may be occurring in affected cities and should be detectable.
Requirements. Horizontal: 10 to 50 m; Vertical: 2 cm to 2 m;
Sampling: monthly to annually; Rate of change: months to decades.
Permafrost Conditions
Permafrost areas are especially vulnerable to ecological damage because biological productivity is low, rates of plant replace-ment or regrowth are low, and root density is low, permitting rapid soil erosion. Most areas of permafrost have a lichen covering that serves as a critical food resource near the base of the food chain. Lichen cover is virtually the only biological retardant for soil erosion and one that is easily destroyed. Increasing development pressures on these areas is common, often due to underlying mineral wealth. This is especially true in the former Soviet Union where the need for hard currency is driving a development boom. Destruction of the lichen or other sparse plant life can lead to extensive damage to the ecosystem through promotion of soil erosion, loss of primary productivity, and destruction of the matted structure of the existing soil.
Because of its sensitivity to ground cover and soil moisture, SAR imaging can provide temporal sampling of permafrost areas under different weather conditions to investigate seasonal changes as well as the impact of human activities on plants and soil structure in permafrost areas. SAR interferometry may also prove to be useful. The change in dielectric constant of water prevents easy correlation of SAR images across the freezing point of water. However, preliminary results suggest that images can be correlated from one period of freeze or thaw to the next (Zebker at al., 1994c). Further research into this possibility is important.
Requirements. Horizontal: 10 to 30 m; Vertical: 2 to 10 cm;
Sampling: monthly to annually; Rate of change: weeks to decades.
Recommendations:
(1) Questions concerning temporal decorrelation using available TOPSAR, ERS-1, JERS-1, and SIR-C data need to be addressed. The critical ques-tion for a dedicated satellite mission for surface change detection concerns the geographic and temporal breadth of its applicability. Studies of grass-lands, glaciated areas, and urbanized areas should be included.
(2) The ability to acquire, process, and interpret existing or soon-to-be avail-able data (TOPSAR, ERS, JERS-1,SIR-C, Radarsat) needs to be improved. In particular, an expansion of processing capabilities for SAR data is crucial. This should be accomplished by broadening SAR activi-ties at JPL, and by preparing a portable version of the processing software to enable university investigators and other groups to exploit SAR inter-ferometry. Similarly, Goddard Space Flight Center (GSFC) should be encouraged to continue and expand their efforts to use laser altimetry in combination with SAR to study topography and surface change.
(3) NASA should continue promoting the release of existing Defense Mapping Agency (DMA) Digital Terrain Elevation Data (DTED).
(4) JPL should investigate the technical possibility, trade-offs, and financial implications of achieving 10-m horizontal resolution with a space-based SAR. An increase in resolution from 30 to 10 m will greatly widen the applicability of space-based data.
(5) Experiments combining multispectral SAR, laser altimetry, and ground truth studies should be pursued to produce a better understanding of the TOPSAT mission prior to final design. We urge SIR-C, TOPSAR, ERS-1, JERS-1, and Goddard laser altimetry to coordinate their efforts to achieve this goal.
(6) We encourage the establishment of a science working group to help guide future efforts related to the research goals listed above and the mission goals listed below. This group should include engineers, operational managers, and scientists.
(7) NASA should seriously consider the dual-satellite L-band SAR mission for launch in the 2000-2002 time frame. The dual-satellite scenario presents the greatest opportunity to accomplish the scientific goals presented in this and previous reports.
(8) NASA should investigate a wide range of orbital scenarios to optimize change detection for the extended portion of this mission. These sce-narios should optimize temporal coverage for large sections of the land surface, allow response to specific events detected from SAR or other data, and provide polar coverage.
Previous Section
Next Section
References
Imaging Radar Home Page
Return to Table of Contents
